Bug 842351 - CMap file "Identity-H" missing due to poppler-data change/cleanup
Summary: CMap file "Identity-H" missing due to poppler-data change/cleanup
Alias: None
Product: Fedora
Classification: Fedora
Component: poppler-data
Version: rawhide
Hardware: All
OS: All
Target Milestone: ---
Assignee: Rex Dieter
QA Contact: Fedora Extras Quality Assurance
Depends On:
Blocks: 630632
TreeView+ depends on / blocked
Reported: 2012-07-23 14:39 UTC by Robert Scheck
Modified: 2012-09-25 06:22 UTC (History)
4 users (show)

Fixed In Version:
Doc Type: Bug Fix
Doc Text:
Clone Of:
Last Closed: 2012-09-25 06:22:46 UTC
Type: Bug

Attachments (Terms of Use)
Situation on RHEL 6 and old Fedora (working as expected) (1.38 KB, text/plain)
2012-07-23 14:40 UTC, Robert Scheck
no flags Details
Situation on Fedora 16/17 (broken, doesn't work anymore) (1.88 KB, text/plain)
2012-07-23 14:40 UTC, Robert Scheck
no flags Details

Description Robert Scheck 2012-07-23 14:39:13 UTC
Description of problem:
Since the poppler-data change/cleanup the CMap file "Identity-H" is missing
which originally was shipped by the ghostscript package. However, CMap of the
ghostscript package is empty and data seems to be delivered by poppler-data,
but poorly incomplete/broken. See the attached files for examples about how
to reproduce.

Version-Release number of selected component (if applicable):

How reproducible:
Everytime, see above and below - and the attachments.

Actual results:
CMap file "Identity-H" missing due to poppler-data change/cleanup

Expected results:
File "Identity-H" again - like in good old times without the poppler mess...

Comment 1 Robert Scheck 2012-07-23 14:40:03 UTC
Created attachment 599797 [details]
Situation on RHEL 6 and old Fedora (working as expected)

Comment 2 Robert Scheck 2012-07-23 14:40:41 UTC
Created attachment 599798 [details]
Situation on Fedora 16/17 (broken, doesn't work anymore)

Comment 3 Robert Scheck 2012-07-23 14:41:19 UTC
Are these few MB overhead from ghostscript really worth the effort of all
this trouble and debugging and so on? Not really. Please fix...

Comment 4 Robert Scheck 2012-07-23 14:43:12 UTC
I think, even all Identity-* files are even missing?!

Comment 5 Robert Scheck 2012-07-23 15:18:40 UTC
CMaps Identity-H, Identity-UTF16-H, Identity-UTF16-V and Identity-V are

Comment 6 Robert Scheck 2012-07-23 20:02:52 UTC
https://launchpad.net/ubuntu/+source/poppler-data/0.4.0-2 - even Ubuntu is
shipping these CMaps. However the strange upstream seems to dislike them?!

Comment 7 Robert Scheck 2012-07-26 23:04:14 UTC
Marek, ping?

Comment 8 Marek Kašík 2012-07-27 07:13:56 UTC
This should be reported against poppler-data. I'm reassigning this.

Comment 9 Rex Dieter 2012-08-31 14:03:55 UTC
Acknowledged, sorry for my not having time to look into this more  deeply, hopefully, I'll have more time next week.

So plan of attack (in case anyone is willing/able to help):

* compare poppler-data-0.4.5 (latest) with prior releases, back at least to 0.4.0 (or more) and take note of what's gone missing, paying careful attention to Identity-* files referenced in comment #4 and comment #5

* manually re-add missing files in popper-data packaging, test if this helps fix the reported issues

* poke poppler/poppler-data upstream about how/why this files changed and/or went missing.

Comment 10 Rex Dieter 2012-09-15 20:16:04 UTC
OK, so looked inside every poppler-data file going back  to 0.2.1, and none of  them contain any file named Identity-*

Indeed, after looking at 
  * add "Adobe-Identity-0 (Special Purpose)", Identity-H and Identity-V file.
    poppler-data upstream reject to add it, but this debian package include it.

so, looks like the ommision is on purpose.  :(

I've found the sources, and included in poppler-0.4.5-4:

I have a question for ghostscript maintainer:

debian's poppler-data seems to symlink cmaps to /usr/share/ghostscript/.../Resource/CMap, whereas the approach here seemed to be  to include in FONTPATH everything matching:

if you want to keep that approach, then 
should probably be added too (i'm currently adding the Identity- files directly in that directory as how debian is doing it).

(without that change, Robert's testcase still fails)


Comment 11 Rex Dieter 2012-09-15 20:29:08 UTC
Hrm, adding /usr/share/poppler/cMap to fontpath doesn't seem to work. 

linking the stuff under /usr/share/poppler/cMap/ into

yields something a little better,  

GS>/DejaVuLGCSans-Bold-Identity-H findfont
Scanning /usr/share/fonts/default/ghostscript for fonts... 20 files, 1 scanned, 0 new fonts.
Scanning /usr/share/fonts/default/Type1 for fonts... 72 files, 36 scanned, 35 new fonts.
Scanning /usr/share/fonts/dejavu for fonts... 30 files, 30 scanned, 30 new fonts.
Scanning /usr/share/poppler/cMap for fonts... 7 files, 7 scanned, 0 new fonts.
Loading a TT font from /usr/share/fonts/dejavu/DejaVuLGCSans-Bold.ttf to emulate a CID font DejaVuLGCSans-Bold ... Done.
Can't find CMap Identity-UTF16-H building a CIDDecoding resource. 
Error: /undefinedresource in /findfont
Operand stack:
Execution stack:
   %interp_exit   .runexec2   --nostringval--   --nostringval--   --nostringval--   2   %stopped_push   --nostringval--   --nostringval--   %loop_continue   --nostringval--   --nostringval--   false   1   %stopped_push   .runexec2   --nostringval--   --nostringval--   --nostringval--   2   %stopped_push   --nostringval--   1868   1   3   %oparray_pop
Dictionary stack:
   --dict:1168/1684(ro)(G)--   --dict:0/20(G)--   --dict:77/200(L)--
Current allocation mode is local
Last OS error: 2
Current file position is 40

Mind you, the only identity files I found so far were:
Identity-H  Identity-V

Comment 12 Rex Dieter 2012-09-15 20:52:50 UTC
Identity-UTF16-H looks like is included in ghostscript tarball, I guess I could use that one.  After including and symlinking that one into
too, the test case succeeds.  yay.

poppler-0.4.5-5 includes that one too:

So, we're left with getting ghostscript to actually find/use the stuff in
/usr/share/poppler/cMap/ and Why doesn't setting gs FONTPATH=.../usr/share/poppler/cMap/ work?

Comment 13 Rex Dieter 2012-09-15 21:38:24 UTC
re-opened bug #630632 the the "ghostscript (still) not using poppler cmaps" problem.

Comment 14 Fedora Update System 2012-09-15 22:11:42 UTC
poppler-data-0.4.5-5.fc18 has been submitted as an update for Fedora 18.

Comment 15 Fedora Update System 2012-09-16 19:16:39 UTC
Package poppler-data-0.4.5-5.fc18:
* should fix your issue,
* was pushed to the Fedora 18 testing repository,
* should be available at your local mirror within two days.
Update it with:
# su -c 'yum update --enablerepo=updates-testing poppler-data-0.4.5-5.fc18'
as soon as you are able to.
Please go to the following url:
then log in and leave karma (feedback).

Comment 16 Robert Scheck 2012-09-16 23:35:20 UTC
The following files are in poppler-data-0.4.5-5.fc18:

The following file is still *missing* in poppler-data-0.4.5-5.fc18:

Comment 17 Robert Scheck 2012-09-16 23:38:52 UTC
a) All Identity-* files came originally from Ghostscript
b) Can you provide a fixed RPM for Fedora 17 updates (see comment #16 first)

Comment 18 Rex Dieter 2012-09-17 01:09:50 UTC
ghostscript (at least version 9.05 as  shippped with f17) doesn't include Identity-UTF16-V

Comment 19 Rex Dieter 2012-09-17 01:14:31 UTC
So, additional open questions:
* is Identity-UTF16-V really needed?
* if so, where do we get it from?   Looks like ghostscript-8.71 last included it, but since ghostscript no longer does, it seems to imply it's not needed or wanted anymore.

Comment 20 Fedora Update System 2012-09-17 11:46:44 UTC
poppler-data-0.4.5-5.fc17 has been submitted as an update for Fedora 17.

Comment 21 Fedora Update System 2012-09-19 12:16:02 UTC
poppler-data-0.4.5-6.fc16 has been submitted as an update for Fedora 16.

Comment 22 Robert Scheck 2012-09-19 22:13:48 UTC
Looks like Identity-UTF16-V is no longer necessarily required with latest
ghostscript...also fine.

Comment 23 Fedora Update System 2012-09-25 06:22:46 UTC
poppler-data-0.4.5-6.fc18 has been pushed to the Fedora 18 stable repository.  If problems still persist, please make note of it in this bug report.

Note You need to log in before you can comment on or make changes to this bug.